Ductal pancreatic adenocarcinoma is almost uniformly lethal and cannot be effectively detected at early stages. Virtually all cases of pancreatic cancer and preneoplastic ductal hyperplasias contain oncogenic Kras mutations, suggesting its importance in this disease. Animal models of cancer that faithfully recapitulate the cognate human condition afford the opportunity to explore the molecular basis of cancer, develop early detection strategies, and evaluate novel therapies. Although mutant mice predisposed to the development of exocrine pancreatic cancer have been previously described, none develop ductal pancreatic adenocarcinoma resembling the human disease. In order to construct a more relevant murine model of ductal pancreatic cancer, we have engineered a mutant mouse strain that harbors an endogenous, conditionally-expressed, oncogenic K-ras G12D allele. By crossing this strain with mouse strains that express Cre recombinase in the pancreas, we observe preneoplastic pancreatic ductal lesions (PanlN) that strikingly resemble those seen in humans. Here, we will investigate the cellular compartments that are capable of initiating pancreatic cancer, determine the relative importance of the two tumor suppressor genes in the Ink4a/ARF locus with regards to tumor progression, and evaluate the role of Notch signaling and MMP7 function in the development of PanlN and PDA. The availability of a suitable murine model of pancreatic ductal adenocarcinoma will allow a detailed assessment of the molecular and cellular features present in discrete stages of tumorigenesis. Such studies are essentially impossible in human pancreatic cancer specimens since they are invariably detected only at very late stages. Information gained from this murine model should facilitate further understanding of human pancreatic cancer.